Modified Method on Confidence Intervals for Variance in One-Factor Components-of-Variance Model

• 발행기관: 한국자료분석학회
• 저자명: Kwan-Joong Kang
• 간행물 정보: 『Journal of The Korean Data Analysis Society (JKDAS)』Vol.11 No.4, 1793~1807쪽, 전체 15쪽
• 주제분류: 자연과학 > 통계학
• 파일형태: PDF
• 발행일자: 2009.08.30

영문 초록

The confidence intervals for σ²A are given in various forms by many authors in onefactor components-of-variance model y ij = μ+ A i + e ij . In this model, Howe’s(1974) method confidence intervals for σ²A are given by an asymptotic confidence intervals. By the simulation results of this confidence coefficients are not fit exactly to 1 -α/2 confidence coefficients. And this confidence coefficients are close to 1 -α/2 when J is large. In this J is too large to use in the simulation. Thus, we would like to give one of the modified Howe’s(1974) method confidence intervals for σ²A in one-factor components-of-variance model. By the simulation results of this modified confidence coefficients are better than those obtained by using Howe’s(1974) method confidence coefficients, and we can simulate when J is small.

목차

1. Introduction

2. One-Factor Components-of-Variance Model

3. Modified Confidence Intervals

4. Simulation Results

References

키워드

Confidence intervals Modified confidence intervals One-factor componentsof-variance mode Variance components